-
Hi guys.
Now something a bit hard that I know nothing if it is possible to do or not.
Something that would be an enhancement.
Perhaps this is not the right place to put it, however here is what it …
-
It would be useful to run one or more of the JACS benchmark relative transformation systems to get an idea for what kinds of problems we are likely to run into.
-
### 🐛 Describe the bug
Just installed pytorch nightly on my Arc 770 using
```
pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/xpu
```
The…
-
### 🐛 Describe the bug
# Summary
Calling `torch.binomial()` produces an unhelpful and counter-intuitive error message when you pass in tensors with the incorrect dtype.
## Reproduction Steps
```…
-
### 🐛 Describe the bug
The following reproducer seems to take forever to compile on torch 2.4.0, I did not let it run to termination, but the regression is >> 3600 seconds:
```
import torch
im…
-
### 🐛 Describe the bug
```
import torch
a = torch.tensor([1, 2, 3])
b = torch.tensor([4, 5, 6])
torch.vdot(a, b)
a = a.cuda()
b = b.cuda()
torch.vdot(a, b)
```
When I run the above code …
-
### 🐛 Describe the bug
Sometimes, we intend to create layer like `torch.nn.Linear(0, d)`, which contains nothing but a single bias. However, when initialize it, because of weight tensor of shape 0*d,…
-
### 🐛 Describe the bug
By testing with the following code, `angle_pi` and `angle_pi_jit` output different results
````python
import torch
angle = torch.zeros(10,device='cpu').uniform_(-7, -3)
…
-
At some point, we should look into plugging accelerated MD into our repex code. This could be useful for general enhanced sampling.
The current AMD code is built using a CustomIntegrator, so it migh…
-
### 🐛 Describe the bug
When running the below code the python interpreter hangs:
```
from multiprocessing import get_context
from torch import randn
def worker(i):
result = randn(1)
p…